ment of the six last lines is wonderfully enhanced by the light and airy measure of the lines that introduce them. -The anapest is capable, according as it is applied, of two effects extremely different; first it is expressive of ease and familiarity, and accordingly is often used with success both in familiar epistles and in pastoral. The other effect is an expression of hurry, confusion, and precipitation. These two, however different, may be thus accounted for. The first is a consequence of its resemblance to the style of conversation: there are so many particles in our language, such as monosyllabic pronouns, prepositions, conjunctions, and articles, on which the accent never rests, that the short syllables are greatly supernumerary. One consequence of this is, that common chat is with greater case, as I imagine, reduced to this measure, than to any other. The second consequence ariseth purely from its rapidity compared with other measures. This effect it is especially fitted to produce, when it is contrasted with the gravity of the iambic measure, as may be done in the ode; and when the style is a little elevated, so as to be sufficiently distinguished from the style of conversation. All these kinds have been employed with success in the Alexander's Feast, an ode that hath been as much celebrated as perhaps any in our language, and from which I propose to produce some illustrations. The poet on recognizing Jove as the father of his hero, hath used the most regular and perfect iambics

The list'ning crowd admire the lofty sound,
A présent déity' they shout around,

A présent déity' the vaulted roofs rebound.

With rávish'd eárs
The monarch hears,
Assumes the gód

Affects to nód,

And seems to shake the sphéres.

But when he comes to sing the jovial god of wine, he very judiciously changes the measure into the brisk trochaic.

Bácchus éver fáir and young.
Drinking joys did first ordáin.
Bacchus blessings áre a tréasure,
Drinking is a soldier's pléasure.
Rich the treasure,

Sweet the pleasure,

Sweet is pleasure áfter páin.

Again, when he describes his hero as wrought up to madness, and setting fire to the city in a fit of revenge, he with great propriety exhibits this phrenzy in rapid anapests, the effect of which is set off the more strongly by their having a few iambic lines interspersed.

Revénge, revénge, Timótheus críes,

See the fúries aríse!

See the snakes that they réar,

How they híss in their háir,

And the sparkles that flash from their ey'es!
Behold how they tóss their tórches on high,
How they point to Pérsian abódes

And glittering témples of their hostile gods.
The princes applaúd with a fúrious jóy;

And the king seiz'd a flambeau with zeal to destroy

So much for the power of numbers. It may not be amiss now, ere I conclude this topic, to make a few cursory remarks on the imitative powers of the several letters which are the elements of all articulate sounds. And first, soft and delicate sounds are mostly occasioned by an equal mixture of consonants with short and monophthong vowels; the consonants being chiefly those denominated liquids, l, m, n, r, and those among the mutes called slender, p, t, k, or c, and ch when they sound as k; to these add v, also z, and s, when they sound as in the two words Zion and Asia. In like manner the duplication of a consonant sounds more delicately than the combination of different consonants. Thus ammiro is softer than admiro, fatto than facto, atto than apto, and disse than dixe. Secondly, strong and loud sounds are better exhibited by diphthongs and long yowels, those of the mutes called middle, and which comparatively may be termed hard, b, d, g, in both its sounds, and j; especially when these are combined with

liquids which render them more sonorous, without occasioning harshness, as in the words, bombard, thunder, clangour, bludgeon, grumble. Thirdly, to roughness the letter h contributes as well as the gutturals. Such is the Greek X, to which there is no corresponding sound in English, though there is in Spanish and in German; also those of the mutes called aspirates, as f, or ph, and th, in both its sounds *, the double r, and all uncouth combinations. Fourthly, to sharp and cutting sounds the fol lowing letters best contribute, s when it sounds as in mass, c when it has the same sound, ch when it sounds as in chide, x, sh and wh; from the abounding of which letters and combinations amongst us, foreigners are apt to remark I know not what appearance of whistling or hissing in our conversation. Indeed, the word whistle is one whose sound is as expressive of the signification, as perhaps any other word whatever. Fifthly, obscure and tingling sounds are best expressed by the nazals, ng and nk, as in ringing, swinging, twanging, sinking; by the sn, as in snuffle, sneeze, snort; and even by then simply when it follows another liquid or a mute, and when the vowel (if there be a vowel interposed between it and the preceding consonant) is not very audibly pronounced, as in morn, horn, sullen, fallen, bounden, gotten, beholden, holpen.-This sound formerly much abounded in English. It was not only the termination of many of the participles, but also of most plurals both of nouns and of verbs. As a plural termination, if we except a very few nouns, we may say it is now entirely banished, and very much, perhaps too much, disused in participles. The sound is unmusical, and consequently when too frequent, offensive, but may nevertheless have a good effect when used sparingly. Besides, it would be convenient, especially in verse, that we could oftener distinguish the preterite from the participle, than our language permits.

Now, of the five sorts of sound above explained, it

Of these one occurs in the noun breath, the other in the verb breathe. The first is the roughest.

may be remarked by the way, that the first is characteristic of the Italian, the second of the Spanish, the third of the Dutch, and perhaps of most of the Teutonic dialects; the fourth of the English, and the fifth of the French, whose final m and n, when not followed by a vowel, and whose terminations, ent and ant, are much more nazal than the ng and nk of the English. I suspect too, both from their prosody and from their pronunciation, that of all the languages above mentioned, the French is the least capable of that kind of imitation of which I have been speaking. On the other hand, I think, but in this opinion I am not confident, that of all those languages the English is, on the whole, the most capable. There is perhaps no particular excellence of sound in which it is not outdone by one or other of them; the Italian hath doubtless more sweetness, the Spanish more majesty, the German perhaps more bluster but none of them is in this respect so various as the English, and can equal it in all the qualities.

So much for the properties in things that are susceptible of a kind of imitation by language, and the degree in which they are susceptible.

PART II.-In what esteem ought this kind of imitation to be held, and when ought it to be attempted?

It remains now to consider what rank ought to be assigned to this species of beauty, and in what cases it ought to be attempted.

As to the first of these inquiries, from what hath been already said it appears very plain, that the resemblance or analogy which the sound can be made in any case to bear to the sense, is at best, when we consider the matter abstractly, but very remote. Often a beauty of this kind is more the creature of the reader's fancy, than the effect of the writer's ingenuity.

Another observation, which will assist us in determining this question is, that when the other properties of elocution are attained, the absence of this kind of imagery, if I may express it by so strong a term, occasions

no defect at all. We never miss it. We never think of it. Whereas an ambiguous, obscure, improper, languid, or inelegant expression, is quickly discovered by a person of knowledge and taste, and pronounced to be a biemish. Nor is this species of resemblance to be considered as on the same footing with those superior excellencies, the want of which, by reason of their uncommonness, is never censured as a fault, but which, when present, give rise to the highest admiration. On the contrary, not the absence only, but even the attainment of this resemblance, as far as it is attainable, runs more risk of passing unheeded than any other species of beauty in the style. I ought however to except from this, the imitation produced by the different kinds of measure in poetry, which, I acknowledge, is sufficiently observable, and hath a much stronger effect than any other whereof language alone is susceptible. The reason why in other cases it may so readily pass unnoticed, is, that even the richest and most diversified language hath very little power, as hath been shown already, in this particular. It is therefore evident, that if the merit of every kind of rhetorical excellence is to be ascertained by the effect, and I know of no other standard, to this species we can only assign with justice the very lowest rank. It ought consequently ever to give place to the other virtues and ornaments of elocution, and not they to it.

As to the other question, In what cases it may be proper to aim at the similitude in sound of which I have been treating; those cases will appear to one who attentively considers what hath been already advanced on the subject, to be comparatively few. Hardly any compositions in prose, unless those whose end is to persuade, and which aim at a certain vehemence in style and sentiment, give access to exemplify this resemblance. And even in poetry it is only the most pathetic passages, and the descriptive parts, to which the beauty whereof I am speaking seems naturally adapted. The critical style, the argumentative, and the didactic, by no means suit it. Yet it may be said, that some of the examples above quoted for the illustration of this subject, are taken